2009 DOE Chemical Hydrogen Storage Center of Excellence

Low-Cost Precursors to Novel Hydrogen Storage Materials

Project ID# ST_20_Linehan

S. Linehan, N. Allen, R. Butterick, A. Chin, L. Klawiter, F. Lipiecki, S. Nadeau, S. November

Rohm and Haas CompanyMay 21, 2009

This presentation does not contain any proprietary, confidential, or otherwise restricted information

2

Project Overview

• Start: March 1, 2005• End: March 31, 2010• Percent complete: 80 %

• System cost• Regeneration processes

– Cost– Energy efficiency– Environmental impacts

Timeline

Partners

Barriers

BudgetTotal

FundingFY08

ActualFY09

Budget*

DOE $3,438K $642K $1,300K

$557K

Phase 2 DOE:ROH Split 70:30Does not include DOE funding to INL ($700K) in Phase 2

ROH $1,524K $275K

3

Objectives/Relevance• Identify cost and energy efficient pathways to “first fill” and

regeneration for ammonia borane (AB) and other borane materials (Phase 2) – Continue experimentation leading to selection of single pathway for low-cost

NaBH4 and further AB process technology development– Guide selection of a top AB regeneration scheme for experimental studies on most

promising alternatives

• Low cost AB (and other borane-based materials) requires low cost NaBH4 for initial system fill– NaBH4 is dominant component to AB costs– Lower cost NaBH4 technologies needed

• Low cost NaBH4 also essential to Metal Hydride Center success

4

Approach/Milestones – AB Cost and Energy Efficiency Estimation

EquipmentList

Material Balance

SizedEquipment

Physical propertiesHeuristics

CapitalInvestment

Aspen IPETM H2A

H2A

RM Prices

ProcessEfficiency

FCHTool

LaborRequirements

Wage Rates

Energy Balance

CapitalCost

MaintenanceProperty Overhead

Labor/RelatedCosts

Raw Material Costs

Energy/Utility Costs

Conceptual

Process Flowsheets

H2A DeliveredH2 Cost

• Methodology developed for determining energy efficiency and delivered costs • Results reviewed with TIAX; feedback incorporated• Baseline cost estimates developed February 2009 (1st fill AB and AB regen

milestone reports)

5

Approach/Milestones – Low Cost NaBH4 for 1st Fill ABIdentify Leading

Pathways

Develop screening and evaluation criteria specific to NaBH4regeneration cycles

Review prior technical and patent literature

Select leading NaBH4regeneration pathways based on theoretical energy efficiencies from reaction energetics and relevant metrics

Demonstrate key chemical and process steps in laboratory studies

Develop flow sheets and preliminary energy requirements and cost estimates for leading systems

Establish complete material balance to determine intermediates and purification requirements

Demonstrate all chemical and process steps

Investigate scalability

Determine Feasibility of

Leading Pathways

Develop single NaBH4 process

Update economics

Detail Performance to

Select Single Pathway

Develop Single Pathway

Sept 2007Go/No Go

Decision for NaBH4 as a

storage material

July 2009milestone

Phase 2

6

Metathesis of ammonium salt and MBH4 in organic solventsnNaBH4 + (NH4)nX → nNH3BH3 + NanX + nH2

Purdue: Ramachandran et al, Inorg. Chem, 2007, 46, 7810-7817

NaBH4 + ½ (NH4)2SO4 NH3BH3 + ½ Na2SO4 + H2 (>95%)

NaBH4 + (NH4)HCO2 NH3BH3 + NaHCO2 + H2 (>95%)

PNNL: Heldebrant et al., Energy & Envir. Science, 2008, 1, 156-16

NH4Cl + NaBH4 NH4BH4 NH3BH3 (up to 99% yield)

Base displacement of borane complexes with ammoniaL-BH3 + NH3 → NH3BH3 + L

Ohio State: Shore et al, WO2007/120511 A2

NH3

-NaClTHF-H2

THF

dioxane

1st Fill AB: Leading Chemistries Identified

Current analysis

7

H2

1st Fill AB: Separation Requirements Determine Feasibility

Reactor Separator

Insolubles Processing

Evaporator Dryer

Recycled SolventFresh Solvent (makeup)

Fresh (NH4)nX

Fresh NaBH4

NanX

Waste Solution to Treatment Facility

Recycled (NH4)nX

Ammonia Borane

Water

Solvent Recovery

Byproducts

Wet ProductDissolved Product

(Water, (NH4)nX)

(NH4)HCO2/dioxane chosen over (NH4)2SO4/THF - easier NH4X / NaX separation- lower solvent usage

Metathesis – NH4 salt

8

1st Fill AB Cost: NaBH4 is Dominant Component

$2.00 $4.00 $6.00 $8.00 $10.00 $12.00 $14.00 $16.00

SBH Price

Amm. Form. Price

Excess Amm. Form.

Production Rate

Sod. Form. Price

Capital Cost

Yield

# Operators

$ / kg AB

$5 / kg$2 / kg $10 / kg

$0.20 / kg $1.00 / kg

5000 MTA20000 MTA

$0.10 / kg$0.50 / kg

$35MM $59MM

16 32

BASE CASE: $9.00 / kg AB

0% E

100%

99% 92%

Sensitivity Analysis

• Cost = $9/kg AB– ~80% from raw materials– NaBH4 at $5/kg accounts for 95% of RMs; 75%

overall• AB cost would be $55-85/kg at current $40-$60/kg

NaBH4• Low cost NaBH4 technologies under development

– 2007 study indicates $1-2/kg NaBH4 at high volume production in regen plants

– Use $5/kg NaBH4 as initial estimate for 1st fill AB scales

AB plant capacity = 10,000 MTAMetathesis – NH4 formate / dioxane

Raw Materials

UtilitiesLabor

Capital

Royalty

0

1

2

3

4

5

6

7

8

9

10

$5/kg SBH

AB

1st

Fill

Cos

t, $

/ kg

AB

$0.15 0.87

0.51 0.32

7.11

9

$2/kWh or $66/kg H2

$4/kWh or $133/kg H2

0

20

40

60

80

100

120

140

160

180

200

0 3 6 9 12 15 18 21 24 27 30AB 1st Fill Cost, $ / kg AB

On-

Boa

rd H

2 Sto

rage

Sys

tem

Cos

t, $/

kg H

2

DOE 2010 Target

DOE 2015 Target

Purdue @ $5 / kg NaBH4

2.5 mol H2 released / mol AB

2 mol H2 released / mol AB

DOE Storage System Targets Require Low-Cost H2 Storage Media

10

Two R&D Approaches to Low Cost NaBH4 Under Development

Pathway ChemistrySchlesinger

(current)4NaH + B(OCH3)3 → NaBH4 + 3NaOCH3

- 25% utilization of Na metal

1-step: NaBO2 + 2x/y M + 2H2 → NaBH4 + 2/y MxOy

2-step: 2x/y M + 2H2 → 2x/y MH2y/x

NaBO2 + 2x/y MH2y/x → NaBH4 + 2/y MxOy - lower-cost metal and lower usage vs. Na- reactive milling

NaBO2 + 2CH4 → NaBH4 + 2CO + 2H2- methane instead of metal reductant- syn gas (CO/H2) byproduct- high temperature to convert B-O to B-H

Metal Reduction

Carbothermal Reduction

Objective: Select single pathway for low-cost NaBH4 in 3Q2009

11

Lab material balance analysis points to recyclable process.

Full accountability for all products.

No intractable byproducts identified.

Isolation and purification steps identified.

Alane work may be relevant to Metal Hydride Center.

Chemical Pathway Established for Metal ReductionFocus on 2-step process via metal hydride intermediate- Offers advantage of higher yields and lower reaction severities

Quantitative recovery of L

88% Mass Balance88% Mass Balance

89% NaBH4 yield99% purity

12

Metal Reduction – Key Chemistry for Reactive Milling Confirmed

3 NaBO2 + 4 AlH3·L 3 NaBH4 + 2 Al2O3 + 4 L

Borate source • NaBO2 : 89% NaBH4, 5% BH• Na2B4O7 + NaX: 62% NaBH4

Alane adduct • L1 : 89% NaBH4, 5% BH• L2 : 1% NaBH4, 29% BH• L3: 2% NaBH4

• 1.1 eq. AlH3-L: 60% NaBH4, 1% BH• 2.0 eq. AlH3-L: 77% NaBH4, 1% BH

• Borate + AlH3-L 29% NaBH4 + 60% BH 89% boron conversion

Stoichiometry

Alternative Approaches

Slurry-based systems

Chemistry(Ball Milling)

Results based on 11B NMR analysisBH = borane intermediate

89% NaBH4 yield (11B NMR)

13

Reactive Milling Modeling Guides Lab Studies, Defines Scalable Process

• Discrete-element modeling provides fundamental understanding of mill motions and particle collisions

• Obtain insight into mill operation, scaleup, and design

• Model also used to help guide lab studies

Lab Mill 1 Lab Mill 2

14

Carbothermal Reduction of Borate – Prior Results Still to be Validated

• Experimental - NaBH4 formation remains elusive – NaBH4 has not been produced but some water-reactive material has

(possibly boranes or Na/NaH)– Equipment issues - exact repetition of prior positive NaBH4 studies difficult – Analytical challenges – Working closely to assist analysis and understand various aspects of

previous results

Established To Be Defined• Collaboration with INL• Prior INL studies: nearly

50% NaBH4 yields for NaBO2 reduction by reducing gases in plasma arc

• High reaction temperatures (>1200°C) dictated by thermodynamics

• Plasma arc process technology commercially viable

• Formation of intermediates and byproducts and required separation

• Reaction quench and heat integration needs

• Best mode for carbothermal reaction (temp sensitivity)

• Scaleup options for commercial high temperature operations

NaBO2 + 2CH4 NaBH4 + 2CO + 2H2

15

ΔH,rxn, 25Ckcal/mol AB

Reactor 1:Digestion (1a) BNH + 1.5 C6H4(SH)2 → 0.5 HB(C6H4S2)·NH3 + 0.5 (NH4)B(C6H4S2)2 5.7

Side Reactions: (1b) C6H4(SH)2 + HB(C6H4S2)·NH3 → (NH4)B(C6H4S2)2 + H2 20.9

Reactor 2: Reduction: (2a) 0.5 (NH4)B(C6H4S2)2 + 0.5 Bu3SnH → 0.5 HB(C6H4S2)·NH3 + 0.5 (C6H4)(SH)(SSnBu3) -15.7Amine Exchange: (2b) HB(C6H4S2)·NH3 + Et2NH → HB(C6H4S2)·NHEt2 + NH3 (g) 1.4Reduction: (2c) HB(C6H4S2)·NHEt2 + 2 Bu3SnH → Et2NHBH3 + C6H4(SSnBu3)2 6.5

Net: (2d) 0.5 HB(C6H4S2)·NH3 + 0.5 (NH4)B(C6H4S2)2 + 2.5 Bu3SnH + Et2NH → -7.8Et2NHBH3 + NH3 + 0.5 (C6H4)(SH)(SSnBu3) + C6H4(SSnBu3)2

Side Reactions: (2e) C6H4(SH)2 + Bu3SnH → C6H4(SH)(SSnBu3) + H2 -10.5(2f) C6H4(SH)(SSnBu3) + Bu3SnH → C6H4(SSnBu3)2 + H2 - 4.0

Reactor 3Ammoniation: (3) Et2NHBH3 + NH3 (l) H3NBH3 + Et2NH -4.1

Reactor 4:Metal Recycle: (4a) 0.5 C6H4(SH)(SSnBu3) + 0.5 H2 → 0.5 C6H4(SH)2 + 0.5 Bu3SnH 5.3

(4b) C6H4(SSnBu3)2 + 2H2 → C6H4(SH)2 + 2 Bu3SnH 14.5

Overall ΔHrxn for target reactions (no side reactions) and no heat recovery of Rxr 2 and 4 product = 25 kcal/mol AB

= 21 MJ/kg H2 (potential for very high energy efficiency)

LANL AB Regen: Heats of Reaction Favorable for High Efficiency Process

demonstrated

16

LANL Spent AB Regeneration Route: Conceptual Process Flowsheet Developed

Plant Design Basis• 225,000 MTA ammonia borane production

• Equivalent to 100 mt/day H2 @ 90% on-stream and 2.5 mol H2 release per mol AB• Negligible losses during fuel shipping and storage• AB not used to generate heat for on-board H2 release • 250 mt/day H2 production as sensitivity

• Spent AB fuel delivered to plant and regenerated AB bulk powder leaves plant• Delivery costs to auto not included since design of AB fuel is not yet defined

17

Capital Recovery and Utilities Dominate LANL AB Regen Cost

0

1

2

3

4

5

6

7

8

9

10

100 TPD H290% on-stream 30 days storage

250 TPD H295% on-stream10 days storage

AB

Reg

en C

ost,

$ / k

g H

2

CapitalLaborUtilitiesHydrogen

$2.57

0.46

1.33

2.06

1.50

Electricity

Natural Gas

Electricity

Natural Gas

$1.97

0.24

1.28

2.06

1.50

AB Production, TPD H2 equivalent 100 250

Baseline AB Regen Plant Cost, $/kg H2

7.9 7.0

Total Investment, $M 520 1000Capital 460 850First Fill Chemicals

(dominated by tin)60 150

Capital recovery plus utilities account for 75% of regen cost

- necessitated by high mass flows and separation requirements

18

Nearly 60% of Capital Cost and 90% of Utility Cost Related to Separations

Opportunities to Lower Cost/Energy Usage• Reduce mass flow• Minimize refrigeration - use alternative reagents to improve thermal stability• Simplify separation scheme - eliminate unessential separations • Reduce separation severity• Consider alternatives to high energy distillation (e.g., crystallization)• Improve heat integration

100 mt/day H2 plant

Other Heating4%

Other Electricity7%

HYDROGEN:22%

Metal Reduction/Amine Exchange Separation

30%

Cooling Tower Water2%

Coolant25%

Metal Recovery Separation

10%

ELECTRICITY: 34%

267 MJ/kg primary energy86 MJ/kg process

HEATING: 44%

349 MJ/kg primary energy326 MJ/kg process

176 MJ/kg primary energy120 MJ/kg process

Primary Energy Usage: 792 MJ/ kg H2 (15% efficiency)616 MJ/kg H2 used in regen plant - excluding H2 feed

Installed Equipment Cost Breakdown

Storage, 8%

Metal Recovery,

16%

Utilities, 20%

Ammoniation, 9%

Digestion, 10%

Metal Red, 38%

(18% reactor20% sep'n)(1% reactor

8% sep'n)

(5% reactor11% sep'n)

(2% boiler3% CTW15% refrig)

19

Collaboration – Essential for Sound Process Development and Analysis Work• AB regeneration processes

– LANL, PNNL, UPenn: Experimental results input– U Alabama: Thermochemical calculations– Rohm and Haas guides Center development work

• First fill AB process analysis– PNNL: Experimental results input

• Low cost NaBH4 for 1st fill AB– INL (sub-contractor): Carbothermal studies

20

• AB regeneration– Investigate options to reduce cost of LANL AB regen scheme

• Reduce mass flow• Utilize less energy intensive separation schemes

– Conduct baseline cost estimate for PNNL, hybrid AB regen routes

• First fill AB– Conduct baseline cost estimate for PNNL and Shore schemes:

• NaBH4 + NH4Cl → NH4BH4 + NaCl → NH3BH3 + NaCl + H2• L-BH3 + NH3 → NH3BH3 + L

– Refine cost estimates with updated NaBH4 cost

Future Work

21

• Low cost NaBH4 for 1st fill AB: continue R&D to identify high-yield, low-cost scalable process– Metal reduction: select best scalable process

• Demonstrate alane formation• Define chemistry and process window • Identify byproduct formation, establish material balance • Develop separation and purification needs• Detail conceptual process and costs

– Carbothermal: progress experimental program to validate prior positive results• Define process window, quench requirements, separation/purification

needs• Detail conceptual process and costs

– Select single top pathway: metal reduction or carbothermal• Continue R&D to define and develop process• Confirm scalability• Update flowsheets and economics• Develop life cycle impacts

Future Work (continued)

22

• AB First Fill– Low cost NaBH4 is dominant factor for producing 1st fill AB at cost required to

meet 2010/2015 DOE hydrogen storage system cost targets. – At $5/kg NaBH4 price using new technology, the metathesis reaction of NaBH4

with NH4 formate developed by Purdue University can produce AB on a commercial scale at about $9/kg, which may meet DOE targets.

– AB synthesis paths differ in separation requirements and could impact AB purity and performance.

– PNNL and Shore routes are undergoing investigation.

• Low-Cost NaBH4 for 1st Fill AB– Alane reduction of borate affords high-purity NaBH4 with high yields.– Chemical pathway has been elucidated: good material balance with no

intractable byproducts observed; points to recyclable process.– Carbothermal reduction of borate to borohydride remains elusive.

• AB Regeneration– Baseline cost estimate to regenerate AB using the thiol-based LANL route:

$7 - 8/kg H2 (exit regen plant) for 100 – 250 MTD H2 equivalent.– Capital recovery and utilities are dominant components of cost, due to high

volume process flows and their separation requirements.– Opportunities are identified to lower processing and separation costs.

Summary

23

Supplemental Slides

26

0123456789

10

1 2

NaB

H4

Cos

t, $/

kg

Metal ReductionCarbothermal

NaBH4 Plant : NaBH4 Regen Plant For 1st Fill AB Capacity, MTA : 173,000 13,000 (100 mt/d H2) (10,000 MTA AB)

$0.2 - 2.3 1

$ 2 - 7

Development of Low-Cost NaBH4Important at All Production Scales

Cost ranges reflect sensitivities in yield, production volume, capital investment, utility costs, byproduct values, and labor costs1 Linehan et al, 2008 AMR : Carbothermal reduction = $2-7/kg H2, metal reduction = $6-12/kg H2